The present invention relates to a data processing apparatus and method, and more specifically, to a data processing apparatus which has a plurality of nodes, each of which includes one or more processors and a main storage device and which is configured to clusters, and to a data processing method implementing said apparatus.
In a conventional data processing apparatus, a plurality of clusters, each of which has one or more CPUs and a main storage device, are interconnected through a network. Clusters process data concurrently. An example of such a conventional data processing apparatuses is disclosed in Japanese Patent Laid-Open No. Hei. 6-231033.
In the conventional data processing apparatus, each cluster is connected through a network, and the communications between the clusters are performed using an extended memory. However, this creates a problem because the size of the entire system increases and the communication speed between the clusters decreases.
In addition, once an apparatus is designed, a designed cluster configuration cannot be easily changed in the conventional data processing apparatus. This creates a problem because few configuration changes are allowed when a cluster configuration is changed, depending on the purpose of the operation of a system.